Nbr

Comment by reviewer

Response

1

The paper presents an analysis of a data set from the Scottish environmental protection agency. This is not a new data set.

As stated in the paper, the original dataset was sourced from SEPA, and as such it is not new. The aim of the paper was to derive metrics from the existing data. We have created a new dataset of annual and absolute maximum rates of rise for the 260 gauging stations which provide the basis for the analysis in the paper. The novelty factor stands in using these created metrics to address the phenomenon of rates of rise and the associated risk to life, a phenomenon rarely addressed by researchers. Given that there are more than 20.000 years of continuously measured 15-min data across the 260 stations, these metrics were a necessity to summarize the issue and derive the conclusions.

2

The analysis are very limited and naïve. There are some basic flaws. The novelty and scholarship are thin.

The reviewer does not address the purpose of the paper which is to highlight the risk to life of very rapid rates of rising level and discharge rather than the focus on peak flow typical of flood risk analysis including flash floods.

We accept that the concepts are simple, but they are not naïve. The analysis is novel in that few papers worldwide have addressed this aspect of flood risk (As noted by CC1 and RC1). Hence, the predominance of References by the authors.

By scholarship, we wonder whether the reviewer means complex? We did not intend to address the hydraulic theory of rapid rise and hydraulic shock which is not necessary for this study.

3

There is no discussion of the errors and uncertainties of the analysed data! This is extraordinary and it would be surprising if the information is not provided by the Scottish environmental protection agency.

There are plenty of references of discharge and flow rate "data" which are presented without clear explanation if these were measured discharges or estimated discharges based upon rating curves. I believe that the gauging stations reported water level data only and that some 'rating curves' were applied.

To clarify our use of rated discharge data we have added the following:

‘The data provided by SEPA are based on 15-min measurements of river level and these are with few exceptions converted to flow by rating equations derived from individual discharge measurements at given levels combined with weir equations. SEPA hydrometry team reviews level measurements monthly and rating curves annually. These reviews result in the correction of data artefacts before publishing the timeseries and necessary changes in the rating curves and flow conversions. We further visually inspected all hydrographs of rapid rates of rise and eliminated spurious records following a comprehensive QC procedure (Fileni et al. 2023)’.

We give further examples of published papers which highlight the importance of ‘rated’ data in large sample studies. The use of an ample number of stations and measurements with a duration of several years allows conclusions to be drawn at large spatial and temporal scale for instance:

Xuan Do et al., 2020 (HESS):Studies the trend on extreme flows across 3666 river gauges from 1971 to 2005.

Iliopoulou et al., 2019 (HESS):Study uses 224 rivers with more than 50 years of data each drawing conclusions of seasonal patterns at a continental scale

Slater et al., 2021 (GRL):Uses 10093 gauge records with information from the before the 80s until the 2000s to assess non stationarity of high return periods in flows

In our study, we use 260 stations from Scotland, commonly used in both scientific (Lane et al., 2019, 2022; Lees et al., 2021) and planning and regulation applications in the UK (Wallingford HydroSolution, 2019) The conclusions that we draw would not have been possible without this data.

4

discharge rating curves of streams and rivers are typically developed based upon steady flow conditions and assumptions. But  It is well-known that the rating curve differs at a given site between the rising hydrograph and declining hydrograph, for the same water depths. The differences increases inversely proportional to the duration of the flood event, with major hysteresis during flash floods.

We are well aware of hysteresis in rating curves and the impact of rapidly changing levels on the equivalent discharges. However, the basis for this research is the observation of such rapid rates of rise in level as to cause a risk to life of river users. The impact of hysteresis on discharge assessment is not critical.

However, on the basis of the reviewer’s comment we have added the following proviso on this effect in the Data section.

‘Rating curves for rivers are typically developed based upon steady flow conditions and rarely take account of hysteresis which can be particularly severe during flash floods, with higher discharges for a given level in rapidly rising levels and lower during the recession. However, potential corrections based on hysteresis would not affect a change in the risk to life of river users and have not been attempted.

5

The bibliographic review is poor

-- A number of relevant literature on flash floods were ignored, including UK studies. The list is too long to develop and a very limited number of recent works are listed below.

We agree that there are numerous papers addressing aspects of flash floods, either in river or in surface water. However, as noted above, there are few addressing specifically the impact of observed rates of rise which this research shows that in Scotland can reach nearly 2 m rise in the 15-min observation interval.

We thank the Reviewer for the list of flash flood papers and consider them individually in their relevance to this research.

6

HALFI, E., PAZ, D., STARK, K., YOGEV, U., REID, I., DORMAN, M., and LARONNE, J.B. (2020). "Novel mass‐aggregation‐based calibration of an acoustic method of monitoring bedload flux by infrequent desert flash floods." Earth Surface Processes and Landforms, Vol. 45, No. 14, pp. 3510-3524 (DOI: 10.1002/esp.4988).

The focus of this paper is on calibration of an acoustic method of monitoring bedload flux with accompanying field measurements of level and estimated discharge on a limited number of infrequent desert floods. Field data acquisition for such research differs from the extraction of critical events from an archive of existing level and flow data. We cannot see any relevance to the occurrence of rapid rates of rise in level and discharge and therefore we do not think it is relevant to our paper.

7

HALFI, E., THAPPETA, S.K., JOHNSON, J.P.L., REID, I., and LARONNE, J.B. (2023). "Transient bedload transport during flashflood bores in a desert gravel-bed channel." Water Resources Research, Vol. 59, Paper e2022WR033754, 17 pages (DOI: 10.1029/2022WR033754).

Here, the authors investigate how bedload transport rates change during the passage of natural flash-flood bores. We found this paper very interesting, notably with respect to the measurements of very rapid rise in level at the bore front on a dry channel and the associated rise in surface water slope and bedload transport (and declining afterwards). As above, it refers to field measurements in individual events and it is not clear how the paper is relevant to the characteristics and geographical distribution of rapid rates of rise in level from archived level and flow.

8

KVOCKA, D., FALCONER, R.A., and BRAY, M. (2015). "Appropriate model use for predicting elevations and inundation extent for extreme flood events." Natural Hazards, Vol. 79, pp. 1791-1808 (DOI: 10.1007/s11069-015-1926-0).

This paper discusses appropriate models for assessing peak water levels and inundation extent and concludes that methods including shock capturing are necessary for steep catchments presumably affected by intense rainfall causing flash flooding. The paper does not discuss the modelling of the rate of rise of the wave front which presumably would also require the inclusion of shock capturing. Though interesting, we do not consider the paper relevant to our analysis.

9

KVOCKA, D., AHMADIAN, R., and FALCONER, R.A. (2018). "Predicting Flood Hazard Indices in Torrential or Flashy River Basins and Catchments." Water Resources Management, Vol. 32, pp. 2335-2352 (DOI: 10.1007/s11269-018-1932-6).

Given our primary interest in the risk to river users of rapidly rising flood levels, this is an interesting paper on assessing human stability in floodwaters. We have added reference to it in the paper. (It is of more interest for our next paper in which we map and discuss the characteristics of catchments in which rapid rates of rise (AWFs) occur. This provides alternative methodologies).

10

The Introduction is very poorly developed. It is self-centered around the authors' publications with self-citations after self-citations.

This is true but it is due to the very limited work done on rapid rates of rise by other authors.

11

The text includes some quantitative discharge numbers without explanations if these were measured discharges or estimated discharges based upon rating curves

We have added a to the Data section to explain the source of the discharge numbers, noted in Response 3. We did not include this in our original text as it is standard practice for SEPA)

12

Line 2 suggests 20,000 years of data. Truly amazing, if true, but most likely a typographic mistake.

This is true as an aggregate across 260 stations as explained in Response 1

13

There is no discussion of the errors and uncertainties of the analysed data! This is extraordinary and it would be surprising if the information is not provided by the Scottish environmental protection agency.

-- There are plenty of references of discharge and flow rate "data" which are presented without clear explanation if these were measured discharges or estimated discharges based upon rating curves. I believe that the gauging stations reported water level data only and that some 'rating curves' were applied.

We have added some sentences into the paper on how the data was QCd. Also see above comment in Response 3

Any mention of discharge and flow rate should explicitly state 'measured' or 'rated', with the implicit limitations of the latter

We have put this detail into the data and methods section.

14

On '3.2 Change in velocity': the authors state "Initial velocity before of the arrival of the AWP [...] is likely to be dominated by wave celerity". This is incorrect and untrue.



In most streams and rivers, the initial velocity will be close to the uniform equilibrium velocity, also called 'normal' velocity, derived from momentum considerations. That is, the longitudinal slope of the water surface would be very close to or equal to the bed slope, and the velocity would fullfill the equilibrium between the gravity force component in the flow direction and the boundary shear force resisting the flow motion.

The reference to ‘domination by wave celerity’ does not refer to initial conditions before the arrival of the AWF but during the AWF. We have added text to clarify the distinction.

15

-- On '3.2 Change in velocity': the 'basic equation c=dQ/dA' is the celerity of a monoclinal wave. The monoclinal wave is a mathematical approximation assuming steady flow conditions before and after the flood front. The approximation does not apply to flash flood.

WE did not find this method practical for estimating celerity for an AWF flash flood so we accept  the reviewer’s criticism and have deleted this paragraph.

16

There are plenty of references of discharge "data" which are presented without clear explanation if these were measured discharges or estimated discharges based upon rating curves. I believe that the gauging stations reported water level data only and that some 'rating curves' were applied.

-- the discharge rating curves of streams and rivers are typically developed based upon steady flow conditions and assumptions. But  It is well-known that the rating curve differs at a given site between the rising hydrograph and declining hydrograph, for the same water depths. The differences increases inversely proportional to the duration of the flood event, with major hysteresis during flash floods.

This is now detailed in the data and methods section with an appropriate reflection of the issues of hysteresis etc. with details in Responses 3 and 4.

17

The section must be drastically restructured and rewritten, with removal of flawed data interpretation.

All the sub-sections related to some interpretation of discharges should be removed: That is, sub-sections 4.3, 4.4, 4.5.

We do not agree with this statement – this flow data is standard flow data used in the UK – we have added more explanation of where this comes from in Responses 1 and 3 and in the paper.